JP2002082981A - Profile managing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a profile managing device which can make the profiles of respective subscribers to be appropriate in the site of communications. SOLUTION: In the profile managing device, the associated degree of the subscriber with respect to a subject is operated and it is managed by a subscriber data base 6 based on the number of registration requesting times with respect to the prescribed subject from document data which the subscriber registers to an electronic conference system 1. A subscriber retrieval part 7 retrieves the subscriber associated with the specified subject by referring to the subscriber data base 6 in accordance with a request from a terminal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a profile management device for managing profiles of participants in an electronic communication space such as an electronic bulletin board.

[0002]

2. Description of the Related Art In recent years, with the spread of electronic terminal devices such as personal computers and mobile phones, the number of participants in electronic communication sites such as electronic bulletin board systems (BBS) and electronic news systems such as USNET has increased. For example, a person with knowledge necessary to solve a predetermined problem, a person with a common interest, and the like, and a friendship are established.

[0003] Conventionally, there is a system in which each participant inputs a word that specifies his / her own interest range and views the profile as profile information. Participants can use other participants' profile information to communicate with people who have the necessary knowledge to solve a given problem or to establish a friendship with a person who has a common interest. You can look up and find the right person.

Here, a system for creating profile information based on the appearance frequency information of words extracted from a document created by a user from the viewpoint that it is difficult to obtain appropriate profile information when the profile information is input has been proposed. JP-A-11-66081, "Profile Acquisition System, Information Providing System, Profile Acquisition Method and Medium".

[0005]

However, in the above-mentioned conventional profile acquisition system and the like, the document created by the user on the terminal is used as a reference for extracting the profile. It is not suitable for searching participants in a communication space.

The present invention has been made in view of the above circumstances, and has as its object to provide a profile management apparatus that manages profile information of each participant in a communication field and can appropriately search for a participant. .

[0007]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems of the prior art, the present invention provides a profile management apparatus for managing a profile of a participant who participates in an electronic communication space in which communication history can be managed by topic. And based on the communication history,
It is characterized by including a relevance calculating means for calculating the relevance of each participant to each topic, and an information managing means for managing the calculated relevance as profile information. As a result, the profile of each participant in a communication place is appropriately managed, and this profile can be used later for a predetermined process such as searching for a participant.

Here, it is preferable that the relevance calculating means refers to the participation frequency of each participant for each topic and calculates the relevance in accordance with the distribution of the participation frequency for each topic. As a result, when a subject does not participate very much in one topic but actively participates in another topic, the degree of relevance to the topic actively participating is highly evaluated. Profile information is calculated.

Further, means for analyzing the communication history in the electronic communication field for each topic and calculating the importance of the word in each topic based on the word frequency distribution of each topic, and And means for managing the importance of words. As a result, the importance of the managed word can be provided to a predetermined process for searching for a topic.

[0010] Further, in response to input of information for specifying the field of interest, profile information managed by the information management means is analyzed, and a participant whose relevance to a topic corresponding to the field of interest is a predetermined value or more is searched for. It is also preferable to include a search means.

[0011] Further, a word specifying a field of interest is input using the importance of the word, and the importance of each of the input words is referred to, and at least one topic is determined based on the input word. It is also preferable to include a search unit for searching for at least one participant according to the degree of relevance of the participant to the selected topic.

The present invention for solving the above-mentioned conventional problem is a profile management method, which analyzes a communication history in an electronic communication space for each topic and associates each participant with each topic. Calculating a degree, a step of managing the calculated degree of relevance as profile information, and, upon request, a step of searching for a participant whose degree of relevance to a specific topic is equal to or more than a predetermined value by referring to the profile information. , Are included.

[0013]

Embodiments of the present invention will be described with reference to the drawings. As shown in FIG. 1, the profile management device according to the present embodiment includes an electronic conference system 1 as an electronic communication space, a database unit 2, a word importance calculation unit 3, and a participant importance calculation unit 4 And topic database 5 and participant database 6
, A participant search unit 7 and an interface unit 8. Here, the participant importance calculating section 4 is used for the relevance calculating means of the present invention, the topic database 5 is used for managing the importance of words, the participant database 6 is used for the information managing means of the present invention, and the word importance is used. The calculation unit 3 corresponds to a unit for extracting important words, and the participant search unit 7 corresponds to a search unit of the present invention.

The electronic conference system 1 is an example of an electronic communication space, in which document data requested by a participant to be registered is registered as a communication history. This document data includes the name of the participant who requested registration, the registration time, the content of the document, and the like. In the electronic conference system 1, the document data registered in the system is classified and arranged in units called articles for each topic in a predetermined range. Further, even in this article unit, a predetermined keynote statement and a series of related statements with respect to the keynote statement There is a topic unit called “thread” consisting of The participant selects an article corresponding to the field of interest, and makes a request for registration of document data to a thread of interest in the article. The database unit 2 stores the registered document data together with information specifying the article and thread for which registration has been requested.

The word importance calculation unit 3 includes a database unit 2
For each thread, perform processing to cut out words from document data belonging to that thread, count the number of appearances of each cut out word, and calculate the importance of the word in each thread based on that. Then, the word, the information for specifying the thread in which the word appears, and its importance are associated with each other in a manner shown in FIG.
To accumulate. Note that the process of extracting words can be performed using a widely known morphological analysis technique, and the process of calculating importance from the number of appearances is described in TFIDF theory (G. Salton & C. Buckley, "Term Weighting"). Approaches
in Automatic Text Retrieval ", Department of Compu
ter Science, Cornell University, 87-881, November,
1987). This TFIDF
In theory, the frequency of occurrence of a specific word in a thread of interest is defined as follows using the number of appearances FTi of the word in the thread of interest and the number of types NTi of all words appearing in the thread of interest.

[0016]

[Equation 1] freqTi = log2 (FTi + 1) / log2 (NTi)

Using the total number STall of threads belonging to the article and the number ST of threads including the word, the distribution (distribution degree) idfTi of the word on the thread is defined as follows.

[0018]

[Equation 2] idfTi = log (STall / ST) +1

Based on these, the importance VTi of the word in the thread is calculated by the following equation.

[0020]

[Equation 3] VTi = freqTi × idfTi

As described above, the more frequent words are evaluated higher by freqTi, while words not unevenly distributed in a specific thread (words appearing in many threads) are evaluated lower by idfTi. Therefore, the significance of this product is
Words that are ubiquitous in a particular thread and appear frequently in that thread will be evaluated higher.

The participant importance calculating unit 4 refers to the database unit 2 and, for each thread, the number F of document data registered in the thread by the participant whose importance is calculated.
Pi is counted, and the application frequency freqPi is defined as follows using the number of participants NPi participating in the thread by applying the TFIDF theory described above.

[0023]

[Equation 4] freqPi = log2 (FPi) / log2 (NPi)

The participation distribution (partial distribution of participants) is calculated from the total number SPall of threads belonging to the article and the number Sp of threads in which the participant whose importance is to be calculated has registered document data as follows. Degree) idfPi is defined as follows.

[0025]

[Equation 5] idfPi = log (SPall / Sp) +1

Then, the importance VPi of the participant is calculated as follows.

[0027]

[Equation 6] Vpi = freqPi × idfPi

Then, the degree of importance VPi of the participant is stored in the participant database 6 as a degree of relevance (profile information) of the participant to the thread in a manner as shown in FIG. As a result, a participant who is ubiquitous in a specific thread and who frequently registers document data in the thread is a participant closely related, and the value of the degree of association is increased.

The participant search unit 7 is connected to a terminal device via the interface unit 8, receives a word representing a range of interest of the participant to be searched from the terminal device, and starts processing. By reference, the importance of the input word in each thread is acquired. Then, at least one thread is selected according to the degree of importance, and a participant having a high degree of relevance for the selected thread is acquired with reference to the participant database 6. Then, at least one participant is specified according to the degree of the degree of association, and the specified participant is notified to the terminal device.

In the above description, a case has been described in which processing is performed for all threads. However, for example, when the number of document data registered in a thread is less than a predetermined number (for example, one), a related When the degree or the importance of a word cannot be calculated significantly, it is preferable to ignore the thread.

Further, as described above, the participation unit of a participant may be an article unit or a thread unit, so that the important unit of the word for each registration unit of document data managed hierarchically as described above. The degree and the degree of relevance of the participant may be calculated and managed, and may be used for a predetermined process such as a search.

Further, here, it is assumed that the participant is a person who actively speaks or writes in a place of electronic communication, but a person who simply browses is also counted as a participant. It is also suitable. As a result, a profile of a person who is interested in a topic in a specific field even if he is not actively participating is appropriately generated.

[0033]

According to the present invention, the means for calculating the degree of relevance analyzes the communication history in the electronic communication space for each topic, calculates the degree of relevance of each participant to each topic, and manages it as profile information. In addition, using the managed profile information, a search process is performed for a participant closely related to a predetermined field of interest. As described above, since the profile information is calculated based on the history in the communication place, the profile of the participant in the communication place can be managed and appropriate profile information can be obtained.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram of a profile management device according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram illustrating an example of data stored in a topic database, in which importance of each word is set.

FIG. 3 is an explanatory diagram illustrating an example of profile information stored in a participant database.

[Explanation of symbols]

1 electronic conference system, 2 database section, 3 word importance calculation section, 4 participant importance calculation section, 5 topic database, 6 participant database, 7 participant search section, 8 interface section.

Claims (8)

[Claims] 1. A profile management device for managing a profile of a participant who participates in an electronic communication space in which a communication history can be managed for each topic, based on the communication history. A profile management device, comprising: a relevance calculating unit that calculates a relevance; and an information management unit that manages the calculated relevance as profile information. 2. The profile management device according to claim 1, wherein the relevance calculating unit refers to a participation frequency of each participant for each topic and calculates a relevance according to a distribution of the participation frequency for each topic. A profile management device characterized by the above-mentioned. 3. The profile management apparatus according to claim 1, further comprising: analyzing a communication history in the electronic communication field for each topic; and analyzing each topic based on an appearance frequency distribution of words for each topic. A profile management device comprising: means for calculating the importance of a word in the above; and means for managing the importance of the calculated word for each topic. 4. The profile management device according to claim 1, further comprising: receiving information for specifying a field of interest, analyzing profile information managed by said information management means, and A profile management device including a search unit for searching for a participant whose relevance to a topic corresponding to a field is equal to or greater than a predetermined value. 5. The profile management apparatus according to claim 3, further comprising: receiving a word specifying a field of interest, referring to the importance of the input word for each topic, and determining at least A profile management device comprising: a search unit that selects one topic and searches for at least one participant according to the degree of relevance of the participant to the topic. 6. A step of analyzing a communication history in an electronic communication space for each topic and calculating a degree of relevance of each participant to each topic; a step of managing the calculated degree of relevance as profile information; Searching for a participant whose relevance to a specific topic is equal to or greater than a predetermined value with reference to the profile information. 7. A module for analyzing a communication history in an electronic communication field for each topic and calculating a relevance of each participant for each topic, and a module for managing the calculated relevance as profile information. A computer-readable recording medium storing a profile management program including the program. 8. A module for analyzing a communication history in an electronic communication space for each topic and calculating a relevance of each participant for each topic, a module for managing the calculated relevance as profile information, And a module for searching for a participant whose relevance to a specific topic is equal to or greater than a predetermined value with reference to the profile information.